摘要
根据城市交叉口交通流的特点,给出了一种交叉口多相位自适应控制算法,综合考虑相邻车道上的车队长度,利用多层BP神经网络实现了道路交叉口多相位模糊控制.仿真结果表明,文中所设计的模糊神经网络控制器能有效地减少单交叉口平均车辆延误,具有较强的学习和泛化能力,是实现交通系统智能控制的一条新途径.
According to the features of traffic flow in urban intersections, a multi-phase self-adaptive control algorithm was proposed. Multi-layer BP neural network was used to realize the multi-phase fuzzy control in road intersections by taking the queue length on contiguous phase lanes into account. Simulation results show that, as a new method of the intelligent control of traffic system, the proposed fuzzy neural network controller can decrease the average vehicle delay in single intersections and it possesses excellent learning and generating abilities.
出处
《华南理工大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2004年第6期67-70,79,共5页
Journal of South China University of Technology(Natural Science Edition)
基金
国家自然科学基金资助项目(60064001)
广东省自然科学基金资助项目(20011707)
关键词
交通控制
模糊控制
神经网络
BP学习算法
车辆平均延误
traffic control
fuzzy control
neutral network
BP learning algorithm
average vehicle delay